Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

Midjourney V6 epic upgrade, netizens exclaimed too lifelike! The magic map has been exposed one after another, and it is said that it will be online today and tomorrow.

2024-02-28 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > IT Information >

Share

Shulou(Shulou.com)12/24 Report--

Xin Zhiyuan reports

Editor: Aeneas is sleepy

[guide to Xin Zhiyuan] Midjourney V6 has just opened its second community evaluation, and netizens are shocked to aphasia by various magical images. The effect of light and shadow is unnatural, the texture of skin and hair is delicate, and the difference between images and photos has completely disappeared!

Midjourney V6, in emergency RLHF. The countdown to the launch has begun!

This morning, Midjourney V6 opened its second community review. Users need to mark 200 pieces and send them for 1 hour.

Netizens have thrown out the magic map one after another, the details of the pictures are lifelike, and the effect is amazing: Zhang Zhang is a blockbuster.

How to evaluate the effect of the V6 drawing this time? to put it simply-the difference between the image and the photo has disappeared.

Sure enough, only today's Midjourney can surpass yesterday's Midjourney.

These magical maps prove once again that no one can shake the position of AI in the field of Midjourney biographies.

You know, the release of Midjourney V5 is in March 2023, and netizens have been waiting for nine months.

When can we use it?

Good news: Midjourney V6 is in the process of a final round of RLHF. When the style fine-tuning is completed, it will be released before Christmas.

According to official sources, V6 is likely to be online today and tomorrow!

This Christmas present is amazing.

Is this really not a real person? Long Ma's skin texture and facial muscles are very close to the reality of the movie.

The epic enhanced skin texture is amazing.

Let's say the following pictures, which one doesn't look like a real person?

Musk would never have seen an AI picture if he hadn't been riding such a big cat and Musk would have appeared in the bottle.

The Cybertruck rendering is also very realistic.

The effect of light and shadow is also unique, and it is difficult to find unnatural places.

No flaws can be found in the treatment of light and reflection.

A random picture shows the texture of a blockbuster movie.

The blood vessels of the character in the picture, the details of the arm guard, and the texture of the skin are naturally without a sense of AI.

Whether it is Groot or animals, there is no distortion in the details.

The hair and whiskers of cats and dogs are all painted beautifully.

Midjourney V6 also paints food very realistically, adding a lot of detail at the same resolution.

All the colors in the picture are based on Van Gogh's Sunflower.

Draw the natural scenery, but also very realistic.

The painting style of V6 has also been further expanded, and traditional Chinese painting has become better.

Netizen: I am really scared. Some netizens said that I am really a little afraid of Midjourney V6.

To get a picture like this, do you just use my previous prompt style, or do you have to learn a new language?

There's nothing to say. This is real life.

Some people say that we have no reason to pay for the material anymore.

The fidelity and quality are great, and the resolution and DPI are impressive.

V6 has brought us into a whole new field, where ordinary consumers can make expert content.

The new hint structure is one of the best artifacts in AI's life chart tool. The reason why Midjourney is occasionally overshadowed by the light of DALL-E and Adobe Firefly is because of a fatal problem: it is unable to understand natural language cues very well.

After all, there is still a certain threshold for writing a professional prompt.

But now, with the launch of V6, that is likely to change.

According to the analysis of designer "Tatiana Tsiguleva", Midjourney V6 can understand users' prompts more accurately, which means that we no longer need to enter strange phrases or words.

The following is a simple structure based on a typical language framework:

Style purpose: to provide a specific aesthetic or artistic style for the work.

Details to be included: preferred style or historical period.

Theme purpose: to clarify the core focus of the picture.

Details to be included: the characteristics of the main object, such as people, objects, animals, etc., including their appearance, color, etc.

Environmental purpose: to create an environment or background suitable for the theme.

Details to include: location (such as indoor, outdoor, fictional scene), environmental elements (such as natural landscape or city scene), specific time of day and weather conditions.

Purpose of composition: to plan the presentation and perspective of the theme and its elements.

Details to include: viewing angle (such as close-up, panorama, bird's-eye view), shooting angle, and specific composition preferences.

The purpose of light: to create the atmosphere and visual effect of the image.

The details that need to be included: the type of light (bright, dark, natural light), the emotion conveyed (such as cheerfulness, mystery), and the ambient effect.

Purpose of additional information: to enrich the hierarchy and depth of the image.

The details you need to include: secondary elements, roles, animals, and their relationship or location layout to the primary object.

Obviously, this prompt approach similar to DALL-E 3 will make V6 more certain and give us more control over the content of the image we want.

The next step is to rub hands and look forward to the official launch of V6.

Some netizens said they were ready.

Not sleeping, waiting excitedly.

Reference:

Https://twitter.com/op7418/status/1737445886165451227

Https://twitter.com/op7418/status/1737445353002344532

Https://twitter.com/ciguleva/status/1737573077633085584

Https://twitter.com/BeezyManzell/status/1737516209988604289

Https://twitter.com/doganuraldesign/status/1737398934073385346

Https://twitter.com/giffboake/status/1737533263671853462

Https://twitter.com/ai_artworkgen/status/1737588573778432406

Https://twitter.com/dnaijatechguy/status/1737617371454701950

Https://twitter.com/Anima_Labs/status/1737592798390460567

Https://twitter.com/PhotoGarrido/status/1737479363476525377

This article comes from the official account of Wechat: Xin Zhiyuan (ID:AI_era)

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 0

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

IT Information

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report